The BET family of epigenetic regulatory proteins are considered `readers' of chemical marks within histones, and the presence or absence of such modifications typically target genes for repression or transcription, respectively. The recognition function of these proteins is accomplished by two domains, BD1 and BD2 (termed bromodomains), that appear to have distinct biological functions. Recent evidence has demonstrated that BETs tend to be overexpressed in human tumors, and considerable efforts by both academic and pharmaceutical laboratories have sought to develop inhibitors of this class of proteins. Such molecules (BETi), have demonstrated efficacy in animals models, although as with all chemotherapeutic agents, off target toxicity is a major concerns and results in drug dose limitation. These compounds target both the BD1 and BD2 bromodomains. We have hypothesized that inhibition of only one BD is necessary for biological activity: using structure guided, medicinal chemistry approaches, we have synthesized BD2-selective inhibitors. These agents are potent cytotoxins to tumor lines in vitro, but are essentially inactive toward normal untransformed cells. In addition, detailed molecular analyses indicate that the BD2 inhibitors modulate expression of a different subset of genes as compared to the pan BD compound (+)-JQ1. In this proposal, we seek to develop BD1-selective inhibitors using the same approaches as that described above, and to compare the molecular, cellular and antitumor properties of BD1- and BD2-selective BET inhibitors in defined pediatric xenograft models. The specific aims of this application are: 1) To use structure-based drug design to generate BD1-specific inhibitors; 2) to determine the molecular and cellular events associated with exposure of tumor cells to these agents; 3) to evaluate than antitumor activity of the BD1- and BD2-selective inhibitors in a panel of pediatric human tumor xenografts. We anticipate that compounds that target the different BDs will modulate expression of subsets of non- overlapping genes both in vitro and in vivo, and that this data can be used to develop biomarkers for drug exposure. Furthermore, we expect that these molecules will exhibit differential toxicity based upon the expression levels of BET proteins in target organs and that they will be considerably less toxic than currently use pan BD inhibitors. Our studies should lead not only to the development of highly specific chemical probes designed to assess BET function both in vitro and in vivo, but potentially to candidate molecules for clinical drug development.